Rapid modeling of segregation-driven metal-oxide adhesion in high-entropy alloys using macroscopic atom model

TL;DR

This paper introduces an extended macroscopic atom model (MAM) that efficiently predicts metal-oxide adhesion in high-entropy alloys by capturing segregation effects and interfacial phenomena, validated against DFT calculations.

Contribution

The paper develops a composition-consistent extension of MAM for multicomponent alloys, enabling rapid and accurate prediction of segregation energies and adhesion properties.

Findings

MAM predicts segregation hierarchy consistent with DFT.

Stronger segregation at Al2O3 interfaces observed.

Hf and Zr enhance adhesion, S weakens it.

Abstract

Accurate prediction of metal-oxide adhesion in high-entropy alloys (HEAs) is challenging because interfacial segregation, atomic environments, and macroscopic thermodynamic quantities are strongly correlated. Relying solely on first-principles approaches is too expensive for exploring composition, solute concentration, and co-segregation effects. To address this, we extend the macroscopic atom model (MAM) for multicomponent alloys using composition-consistent surface fractions and an interfacial pair-probability formalism that captures deviations from random contact statistics. Applied to CoCrFeNi (AlCoCrFeNi) HEA in contact with Cr2O3 (Al2O3), the model predicts segregation energies and work of separation as continuous functions of composition, reproducing the correct segregation hierarchy of Hf, Y, Zr, and S. The stronger segregation tendency at Al2O3 interfaces, and the non-linear dependence of surface energy and adhesion on solute content and co-segregation is also captured. The results are benchmarked with DFT calculations, which shows consistent trends, particularly the strengthening of adhesion by Hf and Zr through strong metal-oxygen bonding and the weakening effect of S. These results demonstrate that the extended MAM provides a physically interpretable, computationally efficient, and quantitatively predictive framework for screening segregation-controlled adhesion beyond the limits of DFT.